gms | German Medical Science

Cancer registries are essential for conducting randomized trials in cancer control. First, they allow identification and follow-up of large numbers of subjects and hence enhance the feasibility of such studies. By allowing the conduct of such trials the knowledge base of cancer control is strengthened. This is largely due to the fact that data collection through cancer registries is quick and inexpensive. Second, cancer registries improve the validity of the trials by allowing comprehensive identification of cancer cases to be used as either trial subjects (in therapeutic or tertiary prevention trials) or trial end-points (in primary and secondary prevention trials). For instance active participation by the subjects is not necessary which decreases loss of information through drop-out, which can be highly selective. Comparable data quality is essential for avoiding information bias. Third, the information efficiency achievable through cancer registries allows large studies with improved precision. This means that a solid foundation for decision-making can be achieved in a more timely fashion and with a more uniform intervention can be assessed than by pooling data from several studies by means of meta-analysis. Several large trials have been conducted in Finland based on cancer registry data. The ATBC trial with nearly 30,000 participants was the first to show in 1998 that antioxidant vitamins vitamin E and beta carotene do not reduce lung cancer among smokers. The Finnish prostate cancer screening trial encompasses approximately 80,000 men and is alone larger than the PLCO prostate cancer screening trial in the US, but has been conducted with only a fraction of the cost. Also, an individually randomized screening trial for colorectal cancer is being conducted in Finland. In the Finnish prostate cancer screening trial, cancer registry is utilized in collection of information on cancer incidence. This is particular important for the control arm of the trial who are not contacted. In addition, information on non-attenders in the screening arm, as well as interval cancers and cases arising after the screening period are effectively identified through the cancer registry. In addition, evaluation of prostate cancer mortality, which is the main end-point of the trial, requires that are deaths among men with a diagnosis of prostate cancer are reviewed by the cause of death committee. Without the cancer registry, the trial would need to be considerably smaller and unable to attain a high degree of coverage of cancer cases.